Remote sensing for mapping ecosystem services to support evaluation of ecological restoration interventions in an arid landscape

This study obtained information through remote sensing of Sentinel-2 imagery, together with soil and terrain data, to estimate the provision of ecosystem services and evaluate integrated ecological restoration interventions in the arid, rural landscape of the Baviaanskloof Hartland Bawarea Conservancy (South Africa), where several integrated interventions have been implemented in areas where decades of small-scale livestock farming have led to extensive land degradation. Interventions included i) long-term livestock exclusion, ii) revegetation of degraded areas, iii) a combination of these two, and iv) essential oil production as alternatives to goat and sheep farming.

Six ecosystem services were assessed: erosion prevention, climate regulation, water flow regulation, fodder provision, biomass for essential oil production and the presence of native species. First, these were estimated by field measurements. Second, the relationships between the quantities of ecosystem services derived from field measurements were analysed with 13 Sentinel-2 indices and four soil and terrain variables. The best model was selected for each service.

Finally, the supply of ecosystem services was compared between disturbed and undisturbed sites. The results showed that models based on Sentinel-2 indices, combined with slope information, can estimate ecosystem service supply in the study area, even when field-based ecosystem service supply levels are low.

The use of Sentinel-2 vegetation indices and ground data to quantify ecosystem services is a first step to improve the monitoring and evaluation of restoration interventions. Results showed that, in the study area, the exclusion of livestock leads to a steady increase in most services.

Main author